This is the logo and if clicked redirect you to home page
Trending News
November 20, 2025 - Third Party

NVIDIA Quarterly Results: A Record Revenue of $57 Billion Driven by AI

NVIDIA Quarterly Results: A Record Revenue of $57 Billion Driven by AI

Introduction: NVIDIA, Driving the AI Revolution

The meteoric growth of NVIDIA shows no signs of slowing down. The California-based company, a pioneer in graphics acceleration and a key architect of artificial intelligence infrastructure, has just announced its historic quarterly results.

With revenue of $57 billion, up 62% year-on-year, and data center revenue reaching $51.2 billion, NVIDIA confirms its position as world leader in the AI economy. The company is now targeting a cumulative revenue horizon of $500 billion by the end of 2026, thanks to the Blackwell and Rubin platforms.

In this article, we will dive into NVIDIA’s figures, strategy, partnerships, and outlook through the following angles:

  • Analysis of financial results and growth drivers
  • Breakdown of the Blackwell architecture and the upcoming Rubin platform
  • Key partnerships with AWS, xAI, and HUMAIN
  • Geopolitical tensions and exposure to the Chinese market
  • Long-term strategic vision and capital allocation

Outstanding Quarterly Results for NVIDIA

Record Revenue at $57 Billion

During the latest earnings presentation, Colette Kress, NVIDIA’s CFO, stated:

"We generated a record revenue of $57 billion, a 62% year-over-year increase, with sequential growth of 22%, an additional $10 billion."

This level of growth is unprecedented in the semiconductor industry and is mainly explained by the insatiable demand for AI infrastructure, fueled by generative language models (LLMs), foundation models, and the rapid spread of AI inference-based applications.

Data Centers: The Main Growth Engine

The Data Center segment alone generated $51.2 billion, a 66% year-on-year increase. This accounts for about 90% of the company's total revenue.

This performance is directly tied to the massive sale of next-generation GPUs — Hopper, Blackwell, but also Ampere, which are still widely used in existing infrastructures.

“The installed GPU base is fully utilized, across all generations.” – Colette Kress

Networking Segment: +162% Growth

The networking division, including products from the Mellanox acquisition, also soared:

  • Revenue: $8.2 billion
  • YoY Growth: +162%

This boom reflects the critical role of bandwidth in modern AI architectures: the bottleneck is no longer just compute, but ultra-high-speed data transfer.

Blackwell & Rubin: The AI Infrastructure of Tomorrow

Blackwell (GB300): Already 2/3 of AI Revenue

The Blackwell GB300 GPU has quickly taken over from the previous generation (GB200), representing two-thirds of AI chip revenue. NVIDIA has managed to accelerate its transition to the new architecture without disrupting supply.

Rubin: Launch Planned for H2 2026

The next platform, called Vera Rubin, is set for a ramp-up in the second half of 2026. It will be based on seven distinct chips and promises, according to Kress, a significant improvement over Blackwell.

“Rubin will offer an ‘x-factor’ in performance, opening a new innovation cycle.”

The stated goal: $500 billion in cumulative revenue between 2025 and 2026 from the Blackwell + Rubin architectures.

Large-Scale Strategic Partnerships

AWS & HUMAIN: 150,000 AI Accelerators Deployed

Amazon Web Services expanded its partnership with NVIDIA via HUMAIN, targeting the deployment of up to 150,000 GB300 units in its data centers.

This is a strong signal that hyperscalers are shifting toward an AI infrastructure specifically designed around the CUDA ecosystem.

xAI: The Bet on a Mega AI Data Center

xAI (Elon Musk’s AI company) and HUMAIN announced the creation of a global network of AI compute centers, including a flagship 500-megawatt site based on NVIDIA’s technology.

“The world is experiencing three major technological transformations, and NVIDIA is addressing all of them simultaneously.” – Jen-Hsun Huang

Tensions with China & Geopolitical Risks

Very Limited H20 Sales in China

The H20 GPU, designed specifically to bypass US sanctions, generated only $50 million in sales this quarter. The significant Chinese orders never materialized, due to:

  • US restrictions
  • The rise of Chinese competitors like Huawei
  • Geopolitical volatility

“Our forecasts currently include no contribution from Chinese data centers.” – Colette Kress

Margins & Capital Allocation: Strategic Discipline

Record Gross Margins: Nearly 75%

  • GAAP Gross Margin: 73.4%
  • Non-GAAP Gross Margin: 73.6%
  • Q4 Guidance: 74.8 to 75%

This places NVIDIA at the top of high-margin tech companies. The company plans to maintain these levels despite rising production costs (notably HBM memory and advanced packaging capacity).

Stock Buybacks & CUDA Ecosystem

Along with the results, Huang confirmed that NVIDIA will continue its stock buybacks while stepping up investments in the software ecosystem (CUDA, AI frameworks, SDKs).

“Every dollar invested expands CUDA’s usage and strengthens our platform.”

Long-term Strategic Vision: Generative AI as a Universal Platform

Inference: The Real-World Use of AI

According to internal data, about 40% of NVIDIA’s current shipments concern inference, which means the production deployment of trained models.

The more this share grows, the more it shows that AI is being integrated into real-world services (search, advertising, translation, chatbots, etc.)

Performance per Watt, Performance per Dollar

One of NVIDIA’s key strengths is its energy efficiency and Total Cost of Ownership (TCO):

“Our architecture delivers the highest performance per watt and per dollar invested.” – Jen-Hsun Huang

Conclusion: NVIDIA Remains the Technology Backbone of the AI Economy

By delivering a record quarter, NVIDIA proves that the AI economy is becoming structural. The company is no longer just selling GPUs; it is structuring data centers, the software ecosystem, industrial partnerships, and the global supply chain.

With visibility of $500 billion in revenue for 2025–2026, an integrated software platform (CUDA), and a global partnership strategy, NVIDIA retains a decisive technological advantage.

Further Reading: Recommended Books & Reports

FAQ – NVIDIA Quarterly Results

1. Why is NVIDIA experiencing such growth?
Thanks to strong global demand for AI infrastructure, especially in data centers, among hyperscalers, and companies developing generative AI models.

2. What does the $500 billion figure represent?
This is NVIDIA’s projected cumulative revenue between early 2025 and late 2026 from its Blackwell and Rubin AI platforms.

3. What’s the difference between Blackwell and Rubin?
Blackwell is the current, already dominant architecture. Rubin is the next-generation platform, planned for late 2026, with significant performance improvement.

4. Why are sales in China weak?
US-imposed geopolitical restrictions prevent the export of advanced chips to China, limiting growth in that market.

5. Are NVIDIA’s margins sustainable?
Yes, despite rising production costs, NVIDIA plans to maintain margins around 75% thanks to sustained demand, scale effects, and optimization of its supply chain.

Sylvain Mouilhaud, Stock Coach

Back to the News